🐻⬇️🏀

2025-2026 NBA Model Performance Analysis

Scope
All Games Ranked Games

All scored games in the selected league and season. AP Poll is excluded here.

Season
2025-2026

Comparing prediction accuracy across 1251 games using multiple rating models.

Model Catalog

7-day holdout coverage: 16/17 models .

Rolling Holdout Curves

Each point is a strict weekly holdout: train on all games before that week, test on that week. This first version uses a 21-day warmup, then 7-day holdouts stepped forward weekly.

Log Loss Brier AUC Accuracy

Weekly strict holdout log loss. Lower is better. Showing 16 models across 26 windows. Click legend items to hide/show series.

Recent Window Winners

Holdout Best Log Loss Runner-up Models
Apr 15 - Apr 19 Adjusted Context Blend 0.566 Home Team Baseline (0.578) 16
Apr 8 - Apr 14 Elo 0.522 Dynamic Bradley-Terry (0.529) 16
Apr 1 - Apr 7 Elo 0.479 Adjusted Context Blend (0.493) 16
Mar 25 - Mar 31 Dynamic Bradley-Terry 0.510 Adjusted Context Blend (0.510) 16
Mar 18 - Mar 24 Elo 0.475 Recency Ensemble (0.502) 16
Mar 11 - Mar 17 Adjusted Context Blend 0.560 Dynamic Bradley-Terry (0.562) 16
Mar 4 - Mar 10 Adjusted Efficiency 0.590 Log Adjusted (0.590) 16
Feb 25 - Mar 3 Log Adjusted 0.495 Adjusted Efficiency (0.495) 16

Model Performance Leaderboard

Models ranked by strict holdout AUC when available (fallback: full-season AUC). Hover over column headers for explanations.

# Model 7d Split AUC Acc Brier LogLoss n AUC 7d Acc 7d Brier 7d n 7d
1 Adjusted Context Blend Adjusted Context Blend Experimental context-heavy win model blending strong team components with rest and venue context. More → STRICT
14g
 - - - - 0 0.800 71.4% 0.193 14
2 Home Team Baseline Home Team Baseline Always favor the home team with a fixed prior. More → STRICT
14g
 0.551 55.1% 0.250 0.693 1100 0.789 78.6% 0.203 14
3 Dynamic Bradley-Terry Dynamic Bradley-Terry Time-evolving paired-comparison model with latent team strength drift. More → STRICT
14g
 - - - - 0 0.756 71.4% 0.201 14
4 Efficiency Efficiency Tempo-adjusted efficiency version of Pythagorean ratings. More → FULL
no 7d
 0.755 68.4% 0.201 0.589 1100 - - - 0
5 Elo Elo Streaming paired-comparison rating with recency baked into sequential updates. More → STRICT
14g
 0.744 68.6% 0.205 0.599 1100 0.733 78.6% 0.196 14
6 Bradley-Terry Recency Bradley-Terry Recency Static Bradley-Terry with exponential recency weights on newer games. More → STRICT
14g
 - - - - 0 0.733 78.6% 0.202 14
7 Core Ensemble Core Ensemble Equal-logit blend of Elo, recency BT, recency margin, log-adjusted pyth, and points off/def. More → STRICT
14g
 - - - - 0 0.733 57.1% 0.213 14
8 Margin Margin Linear team-strength model fit on point differential instead of binary wins. More → STRICT
14g
 0.746 68.0% 0.210 0.609 1100 0.689 57.1% 0.220 14
9 Points Off/Def Points Off/Def Raw points regression with separate offensive and defensive team parameters. More → STRICT
14g
 0.755 67.6% 0.207 0.603 1100 0.689 57.1% 0.220 14
10 Recency Ensemble Recency Ensemble Equal-logit blend of Elo, recency BT, recency margin, log-adjusted pyth, and recency points off/def. More → STRICT
14g
 - - - - 0 0.689 64.3% 0.214 14
11 Margin Recency Margin Recency Margin regression with exponential recency weights on newer games. More → STRICT
14g
 - - - - 0 0.667 57.1% 0.226 14
12 Points Off/Def Recency Points Off/Def Recency Off/def points regression with exponential recency weights. More → STRICT
14g
 - - - - 0 0.667 57.1% 0.225 14
13 Bradley-Terry Bradley-Terry Static logistic paired-comparison model with one team strength parameter. More → STRICT
14g
 0.751 68.5% 0.205 0.597 1100 0.644 57.1% 0.213 14
14 Pythagorean Pythagorean Pythagorean win expectation from raw points scored and allowed. More → STRICT
14g
 0.754 68.5% 0.214 0.619 1100 0.600 57.1% 0.242 14
15 Adjusted Efficiency Adjusted Efficiency Opponent-adjusted efficiency model with separate offensive and defensive components. More → STRICT
14g
 0.753 68.1% 0.203 0.592 1100 0.600 57.1% 0.237 14
16 Log Adjusted Log Adjusted Log-scale adjusted efficiency model that downweights blowout leverage. More → STRICT
14g
 0.753 68.1% 0.203 0.592 1100 0.600 57.1% 0.237 14
17 Avg Margin Baseline Avg Margin Baseline Predict from simple average scoring margin in the training window. More → STRICT
14g
 0.765 70.0% 0.200 0.587 1100 0.600 57.1% 0.237 14

Methodology

ELO / Bradley-Terry

  • ELO: Iterative updates, K=64, HCA=100
  • BT: Static logistic regression on all games
  • Both model win probability, not margin
  • ELO updates after each game; BT fits once

Pythagorean Models

  • Raw: Classic points scored/allowed formula
  • Efficiency: Pace-adjusted (pts per possession)
  • Adjusted: Opponent-adjusted efficiency
  • Log: Log-linear multiplicative scale

Margin Regression

  • Team-level ridge regression on point margin
  • Linear Bradley-Terry (margin target)
  • Alpha=0.05 (CV-tuned)
  • Learns home advantage from data (~6 pts)

Baselines

  • Home Team: Always predict home wins (60%)
  • Avg Margin: Higher average margin wins
  • Models should beat these to add value